In sub-Saharan Africa rates of tuberculosis (TB) and HIV infection are extremely high. TB is the leading cause of death in HIV-infected adults and often is the initial clinical manifestation. TB in HIV-infected patients carries a high risk of death which occurs predominantly in the first days to weeks after TB diagnosis [2-5]. Other severe lung infections (bacterial pneumonias, Pneumocystis, cytomegalovirus) are also major problems in people with advanced HIV disease, although incidence and coinfection rates are poorly delineated. Detailed characterization of HIV-associated lung infections is needed to improve diagnosis, to prevent new infections, and to alter this high early mortality. In settings with limited resources, understanding the interaction between TB, HIV and other respiratory co-infections in increasing morbidity and mortality will help to direct resources efficiently and to allow coinfected patients to survive long enough to access life-sparing antiretroviral (ARV) therapy. We will recruit 750 individuals from an existing fully enrolled, funded, and consented cohort of approximately 1100 patients and an additional 250 individuals using a clinical cohort established with PEPFAR funding. We will gather longitudinal data through the transition toward rapidly evolving access to pre-ARV clinical care and ARVs in Soweto, South Africa. This clinical database in collaboration with the other successful cohort databases will be a rich resource for both clinical epidemiologic studies and the study of basic pathobiologic mechanisms of HIV-associated lung complications. The proposed study is expected to provide information of fundamental importance for public policy and clinical practice, identifying effective and sustainable clinical management strategies to improve local standards of care for patients with HIV infection, to decrease early mortality after HAART and while awaiting HAART, and to identify the key target topics for future clinical trials.